1. Field of the Invention
This invention generally relates to a displacement estimation method and, more particularly, to a displacement estimation method capable of dynamically adjusting the quality threshold of an image frame and a displacement estimation device using the same.
2. Description of the Related Art
An optical displacement estimation device utilizes a plurality of images acquired by an image sensor to estimate a displacement of the displacement estimation device with respect to a working surface. Conventionally, the displacement can be calculated by, for example, the correlation between images.
However, the quality of an image acquired by the image sensor can affect the accuracy of the displacement estimated by using the image. Therefore, conventionally an image acquired by the image sensor will be identified firstly to determine whether the image is suitable for displacement estimation, and the displacement will be estimated by using valid images. A method of identifying an image quality generally includes the steps of: comparing a predetermined fixed threshold and a quality parameter obtained according to the content of an image; and if the quality parameter is higher than the fixed threshold, identifying the image as a valid image that is suitable for displacement calculation.
A low image quality is generally due to a high noise level. When an image acquired by the image sensor is identified as an invalid image, the image quality of the image can conventionally be adjusted by changing a sampling parameter. However, the change of the sampling parameter will also change the noise level of the image. Accordingly, if a fixed quality threshold is utilized, an error displacement may still be outputted due to the use of an invalid image for displacement estimation.
Accordingly, a displacement estimation method that can dynamically adjust the quality threshold according to the sampling parameter or the image noise level is required by the field.